[
https://issues.apache.org/jira/browse/CASSANDRA-8457?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15967513#comment-15967513
]
Jason Brown commented on CASSANDRA-8457:
----------------------------------------
[~aweisberg] and I spoke with Scott Mitchell of the netty team and we have a
potential solution better than what I put in the last comment. The TL;DR is to
add an iterator to netty's {{ChannelOutboundBuffer}} and walk the "queue"
looking for expired items. It requires a patch netty (which I'm working on),
but should be straight forward. More details soon...
> nio MessagingService
> --------------------
>
> Key: CASSANDRA-8457
> URL: https://issues.apache.org/jira/browse/CASSANDRA-8457
> Project: Cassandra
> Issue Type: New Feature
> Reporter: Jonathan Ellis
> Assignee: Jason Brown
> Priority: Minor
> Labels: netty, performance
> Fix For: 4.x
>
>
> Thread-per-peer (actually two each incoming and outbound) is a big
> contributor to context switching, especially for larger clusters. Let's look
> at switching to nio, possibly via Netty.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
